|Publication number||US5850264 A|
|Application number||US 08/575,175|
|Publication date||15 Dec 1998|
|Filing date||19 Dec 1995|
|Priority date||30 Jun 1994|
|Publication number||08575175, 575175, US 5850264 A, US 5850264A, US-A-5850264, US5850264 A, US5850264A|
|Original Assignee||Intel Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (9), Referenced by (16), Classifications (16), Legal Events (5)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application is a continuation of application Ser. No. 08/268,961 filed on Jun. 30, 1994 now abandoned.
This invention relates to digital video transmission systems, and, more particularly, to techniques for accomplishing increased efficiency of digital video image transmissions.
The original analog technique used for video image transmission and display comprised sweeping an amplitude-modulated electron beam across a video screen at a standard repetition rate and doing a fixed number of such sweeps per image frame using a "roster scanning" approach. In order to accomplish such roster scanning on a screen of moderate size (e.g. 14" wide by 12" high) using a standards 525 horizontal lines per frame, while "refreshing" the screen at a rate of 60 frames per second it would have required an excessive frequency bandwidth for the video signal. In order to deal with this problem the technique of "interlacing" was introduced. Using this approach, the even and odd horizontal lines are alternately refreshed as even and odd frames, each at a rate of 30 frames per second. The human eye effectively integrates the overlap in a manner that "tricks" the human viewer into apparently seeing the entire screen being refreshed 60 times per second. Actually, though, the eye is seeing only 30 even frames and 30 odd frames alternately each second. This allows for a reduction of the required video signal bandwidth by a factor of two.
The digital video industry in North America and Japan currently has a standard for "Full Motion Video" of 30 frames per second. This rate gives moderately good results except for fast movement of the subject. In digital video, the entire screen is normally refreshed at this rate, and the frame is not broken up into separate fields.
Applicant has discovered a novel technique for producing a sequence of digital video images at the standard full motion video refresh rate which results in an acceptable moving image on a video screen for a relatively slow-moving subject, while at the same time reducing the amount of digital data that must be processed. The technique employs "pseudo-interlacing" of the digital video data as follows: all even horizontal lines of pixel data of a first full original frame of displayed digital data are combined with all odd horizontal lines of pixel data intended for displaying a second full original frame of pixel data immediately following the first full frame to create a first new pseudo-interlaced frame for processing and subsequent display. A second new pseudo-interlaced frame is produced similarly by combining the above-mentioned odd horizontal lines with the even horizontal lines of pixel data intended for display of a third full original frame. This process is continued in an even-odd, odd-even sequence throughout the entire set of original digital video data.
Use of the above-described pseudo-interlacing reduces by almost half the amount of digital video data that must be processed and transmitted or stored. This due to the fact that for any given full original frame of pixel data only one-half of the data (i.e. either the odd or even lines thereof) are encoded for transmission. Upon reception the data is reconstructed into full frames wherein each frame gets one-half of its data (i.e. either the odd or even lines) from its immediately-preceding frame. The reconstructed data is then displayed as full frames at the full motion video refresh rate. The resulting video image is not significantly degraded as long as the motion of the subject is relatively slow (as, for instance, is normally the case in digital video teleconferencing).
FIG. 1 shows a typical digital video frame.
FIG. 2 depicts the systematic scheme used for discarding partial frames.
FIG. 3(a) shows the preferred pattern for encoding a "half-and half" implementation of the invention.
FIG. 3(b) clarifies one way to synthesize a displayable video image using pseudo-interlacing.
FIG. 4 is a system block diagram of apparatus suitable for implementing the invention.
FIG. 5 depicts a flow chart for implementation of the "discarding" encoding technique of the instant invention.
FIG. 6 depicts a flow chart for implementation of the pseudo-interlacing synthesis technique of this invention.
In a preferred embodiment of the digital video transmission system of the instant invention, each full frame of original digital video data is divided into two separate fields of data as depicted in FIG. 1. The "even" field consists of even-numbered lines 2 through 480, and the "odd" field consists of odd numbered lines 1 through 479. Each line is divided into 640 separate pixels as shown, yielding a total of 307,200 total pixels per frame. Each full-sampled color-encoded YUV pixel requires 24 binary bits of information, yielding a total of 7,372,800 binary bits per frame. At a frame refresh rate of 30 frames per second, this amounts to a mind-boggling real-time processing rate of 221,184,000 binary bits per second| This is prohibitively high, even with the most-advanced Intel Inside® systems currently available. There exist many approximation or "compression" techniques for reducing this rate in practical applications, such as truncation, run-length ("RL") coding, color subsampling, color lookup tables ("CLUT"), and discrete transforms.
The instant invention comprises a novel approach to reduce the above-mentioned real-time processing rate by approximately a factor of two. When used in conjunction with some or all of the currently-available compression techniques, very large reductions in the required processing rate can be accomplished while at the same time maintaining good picture quality. These various techniques, used to extreme, tend to produce distinctive corresponding characteristic "artifacts" (i.e. picture degradations) such as pixelation, contouring, and blurring. But, if each is used in moderation and in combination with the others, they can produce a good quality picture which is substantially free of humanly-visible artifacts. The instant invention, for instance, results in a moving image that is apparently free of artifacts unless the subject moves quickly. A sudden motion by the subject can result in a perceptible jaggedness of non-horizontal transition lines in the video image. But, for a slow-moving subject, no such artifacts are observed. So, at least for certain applications, the technique of this invention provides a major benefit with little or no perceptible tradeoff in image quality.
By referring again to FIG. 1, in conjunction with FIGS. 2, 3(a), and 3(b), basic concept of this invention can be readily gleaned. The odd and even fields of the sequential video frames are alternately digitally encoded. This entails encoding one-half as much pixel data as is available from the original video source. As shown in FIG. 2, half of the originally-available video data is discarded in a regular sequential pattern before the digital encoding is implemented. This concept is also depicted in FIG. 3(a) wherein the information in the dotted-line odd and even half frames is not encoded for transmission. One might be inclined to conclude that such a drastic elimination of information would result in serious degradation of the received image. But, on the contrary, Applicant has discovered that by a suitable synthesis of this incomplete data, a quite acceptable good quality video image can be created (if the subject does not make sudden movements).
Applicant's new-found technique for accomplishing a suitable synthesis of the incomplete data is shown schematically in FIG. 3(b), wherein F1 * corresponds, in time sequence, to the original frame of video data F1 (in FIG. 2), F2 * corresponds to F2, and so on. The thus newly-interleaved (i.e. pseudo-interlaced) frames each have a full frame's worth of digital data, albeit not quite the original data. The human eye has difficulty, however, detecting any significant difference in the resulting displayed image vis-a-vis the image that would have resulted if the entire originally-available video data had been transmitted. It is not important whether the first half frame discarded is an even half frame or odd half frame, as long as from that point on the opposite type of half frame is discarded alternately in sequence. Hence, for example, F1 * could just as well be composed of F12 combined with F21, and F2 * could be composed of F32 combined with F21 and so on. The resulting displayed pseudo-interlaced video image would be substantially indistinguishable from the pseudo-interlaced image previously described above.
It is, of course, possible to carry this idea further by, for example, discarding two out of three lines in ordered sequence from each of the originally-available video data frames, and then synthesizing pseudo-interlaced frames from parts of three sequential originally-available frames in a manner similar to that described above. Or, alternatively, one could discard less than one half of the originally-available data (one-third or one-fourth, for instance) and similarly synthesize frames for display comprised of parts of two sequential originally-available frames. So Applicant's idea of pseudo-interlacing is not limited just to the "half and half" preferred embodiment described originally above, but rather extends to any like arrangement involving the discarding of lines (or columns) or partial-lines (or partial-columns) combined with subsequent systematic synthesis of a full video frame from the remaining data actually transmitted. The usefulness of any particular variation of Applicant's pseudo-interlacing invention herein described will be determined by the particular requirements of the practical application for which it is employed. Appropriate empirical testing may be needed to determine suitabilities for various applications, as will be apparent to those skilled in the digital video art area.
Referring now to FIG. 4, therein depicted is a system block diagram that describes a typical digital video system within which the technique of the instant invention can be used. Therein, a frame buffer 402 receives full-frame digital video data either from a digitizer 401 or an alternate digital video source. Encoding video processor 403 receives sequential frames of digital video data and applies the pseudo-interlacing odd-half-frame-even-half-frame alternate discarding method described previously herein to the digital video data, and then sends the retained data as an encoded bit stream either to storage device 404 or along communications channel 405. The thus-encoded bitstream received at the opposite end of communications channel 405 is then directed to decoding video processor 406 which receives the encoded data and applies the full-frame pseudo-interlacing synthesizing technique described hereinabove to the received bitstream, creating a full-frame pseudo-interlace video data stream for application to the display frame buffer 407. The output of frame buffer 407 is then transmitted to a display or storage device (neither shown) capable of direct reception of sequential digital video frame data, or, alternatively, is transmitted to a digital-to-analog encoder that provides an analog video output for display on an analog display device (not shown).
The encoding video processor 403 implements the alternating digital video data discard process described above and depicted in the flow diagram of FIG. 5. The decoding video processor 406 implements the generation of sequential full frames of pseudo-interlaced digital video data also as described above and depicted in the flow diagram of FIG. 6. Those skilled in the art of digital video processing will recognize that implementing these disclosed processes on any particular video processor is a simple matter of programming. It is to be understood that the instant invention is not limited in scope to the embodiments described above, but rather is defined by the appended claims and fair equivalents thereof.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US3715483 *||11 Dec 1970||6 Feb 1973||Bell Telephone Labor Inc||Bandwidth reduction system for use with video signals|
|US4060832 *||23 Jun 1976||29 Nov 1977||Societe Anonyme De Telecommunications||Digit rate reducing method in video signal transmission|
|US4620225 *||7 Sep 1983||28 Oct 1986||Robert Bosch Gmbh||System for TV transmission|
|US4984082 *||22 Jun 1989||8 Jan 1991||Kabushiki Kaisha Toshiba||Circuit for displaying picture of multiple channels|
|US5072296 *||23 Feb 1990||10 Dec 1991||Massachusetts Institute Of Technology||Video frame reduction/reconstruction method and apparatus|
|US5146325 *||29 Apr 1991||8 Sep 1992||Rca Thomson Licensing Corporation||Video signal decompression apparatus for independently compressed even and odd field data|
|US5174641 *||25 Jul 1990||29 Dec 1992||Massachusetts Institute Of Technology||Video encoding method for television applications|
|US5253058 *||1 Apr 1992||12 Oct 1993||Bell Communications Research, Inc.||Efficient coding scheme for multilevel video transmission|
|US5384598 *||20 Oct 1992||24 Jan 1995||International Business Machines Corporation||System and method for frame differencing video compression and decompression with frame rate scalability|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US6240212 *||23 Mar 1998||29 May 2001||Samsung Electronics Co., Ltd.||Scan interleaving method|
|US6459455||31 Aug 1999||1 Oct 2002||Intel Corporation||Motion adaptive deinterlacing|
|US6507368 *||28 Jan 2000||14 Jan 2003||Canon Kabushiki Kaisha||Control of image signals in liquid crystal displays|
|US6535300 *||24 Jun 1998||18 Mar 2003||Brother Kogyo Kabushiki Kaisha||Picture signal processing apparatus and method|
|US6553153 *||3 Dec 1998||22 Apr 2003||Chips And Technologies, Llc.||Method and apparatus for reducing video data|
|US6897903||31 Aug 2000||24 May 2005||Micron Technology, Inc.||Apparatus for detecting mixed interlaced and progressive original sources in a video sequence|
|US7068847 *||8 Sep 2000||27 Jun 2006||Sunplus Technology Co., Ltd.||High-resolution still picture decoding device|
|US7463306 *||25 Feb 1999||9 Dec 2008||Thomson Licensing||Processing interlaced and pseudo interlaced signals|
|US8159607 *||17 Apr 2012||Sony Corporation||Video signal display system, video signal reproducing apparatus, and video signal display method|
|US8817132 *||20 Jan 2010||26 Aug 2014||Ambarella, Inc.||Hardware partial frame elimination in a sensor interface|
|US20030043913 *||22 Aug 2002||6 Mar 2003||Onno Eerenberg||Videodataprocessor|
|US20040218670 *||2 May 2003||4 Nov 2004||Lai Jimmy Kwok Lap||Method and apparatus for reducing the bandwidth required for transmitting video data for display|
|US20090310017 *||17 Dec 2009||Sony Corporation||Video signal display system, video signal reproducing apparatus, and video signal display method|
|US20140082685 *||30 Apr 2012||20 Mar 2014||Samsung Electronics Co., Ltd.||Method and apparatus for adjusting data transmission rate in a wireless communication system|
|CN101605221B||11 Jun 2009||4 Apr 2012||索尼株式会社||Video signal display system, video signal reproducing apparatus, and video signal display method|
|WO2003019948A1 *||12 Jul 2002||6 Mar 2003||Koninklijke Philips Electronics N.V.||Video data processor|
|U.S. Classification||348/469, 375/E07.254, 348/459, 386/E05.013, 348/448, 375/240.25|
|International Classification||H04N5/44, H04N5/926, H04N7/46|
|Cooperative Classification||H04N19/132, H04N19/587, H04N7/012, H04N5/9261|
|European Classification||H04N7/01G3, H04N5/926B, H04N7/46T2|
|3 Jun 2002||FPAY||Fee payment|
Year of fee payment: 4
|9 Jun 2006||FPAY||Fee payment|
Year of fee payment: 8
|19 Jul 2010||REMI||Maintenance fee reminder mailed|
|14 Oct 2010||FPAY||Fee payment|
Year of fee payment: 12
|14 Oct 2010||SULP||Surcharge for late payment|
Year of fee payment: 11